AI: Sustainability tool or more scope for greenwashing?
AI has huge potential to make businesses more sustainable. It is already being deployed by companies including Google to cool data centres, in hospitality to track and reduce food waste, and by governments such as Indonesia and Peru to show near real-time vessel movements in the ocean to combat illegal fishing.
In 2023, governments and businesses are continuing to strive towards improved sustainability. Environmental, social governance (ESG), is high on boardroom agendas. Businesses want to work with suppliers and partners who have similar values and are equally committed to sustainability. And, less altruistically, ESG is increasingly a focus for scoring when tenders are awarded, or investment is levied.
From a legislative stance, businesses will soon have to comply with the Corporate Sustainability Reporting Directive, which obligates financial market participants to disclose their non-financial and diversity information.
Businesses are therefore actively looking for green solutions that can also improve their marketability and, ultimately, their bottom line. AI is being touted as something that can manage environmental impacts and climate change, whilst also improving business efficiency – a win-win.
Is AI as green as it seems?
When environmental claims are unproven, over-inflated or just incorrect, this is ‘greenwashing’. When implementing AI solutions, there is often little detail given at the micro-level on how AI will save the planet any more effectively or efficiently than traditional computer-human operations.
When implementing AI and measuring the energy-savings it can produce, this needs to be offset against the electricity consumption of AI systems themselves, as this is potentially substantial. It has been calculated that AI’s global carbon footprint might foreseeably be equal to that of the aviation industry.
Until all AI is powered by renewable electricity, including using sustainable data sources, AI’s energy consumption must be taken into account when making claims about the energy-saving capabilities of AI. It is only a matter of time before AI greenwashing undergoes the same scrutiny by private and public auditors and is more heavily legislated against and regulated.
Competition bill looms
We are imminently waiting for the UK Digital Markets, Competition and Consumer Bill, under which it is expected that companies could face fines of up to 10% of their global turnover for breaches of consumer law.
The bill is expected to grant regulators with new powers to use against companies that make misleading environmental claims. This bill, combined with the new EU AI Act, and divergent ‘sector by sector’ approach to regulation in the UK, is going to increase the regulatory considerations for businesses deploying AI technologies.
This is only going to increase, as we see other technologies where AI is crucial, such as the metaverse, develop.
Could AI deliver better ESG standards?
From efficiency gains in agriculture and energy supply to sustainable supply chains and environmental monitoring, the claims made about the green potential of AI are vast.
But there is little published information about how AI use in these fields will be deployed; how AI software will differ from non-AI software; and whether that difference will actually be beneficial in reducing the rate of climate change or meeting biodiversity and ESG goals. Developers and users risk being challenged on such green claims in 2023.
AI is already being used as a tool to combat greenwashing. It is being deployed to analyse a company’s publicly available information on the web to detect early signs of greenwashing or identify related risks.
The hope is this increased scrutiny will force higher ESG standards. There is an irony to being caught out by an AI system for making false statements about the energy efficiency and ESG benefits of the AI system your business has deployed.
So where does this leave businesses in the brave new world of AI, greenwashing and sustainability? Here are some practical steps to keep in mind.
– Marketing statements and representations made by an organisation regarding its ESG commitments and how AI is improving its sustainability must also take into account the AI systems themselves and the energy the AI consumes, not just their application. This means involving legal teams in deploying these technologies and the statements made concerning their sustainability.
– Where AI is being used to enhance ESG standards, proper consideration needs to be given to how the delivery of these tools can be monitored and assessed. As with the deployment of all AI technologies, there should be someone in the business who has an understanding of the ‘output’ and can verify its accuracy and identify potential issues and areas to improve on.
– Customers should review and be prepared to challenge AI and ESG claims made by suppliers to ensure they are accurate. Where overambitious statements have been made, legal advice should be obtained.